Re: [NTLK] Newton 2x00 internal ram

From: Jeff Sheldon (jeff_at_jurai.net)
Date: Wed Oct 16 2002 - 23:30:32 EDT


On Wed, 16 Oct 2002, Matthew B.Hershberger wrote:

> operating system? For a senior college project, I was considering
> porting a scaled down version of linux to the newton, if I can get
> enough information on the basic of the hardware.

If you pull this off, I'll love you forever.

The NewtonOS won me over on my first night with a Newton back (no jokes
please) in '94, but I'd like to have a public depository of how to put
-anything- on the hardware if one chooses.

There's quite a few roadblocks right now due to not enough public material
on specific chips. I've done some guesswork and made a number of phone
calls, but I'm not certain yet as to what the results will be. So unless
you're good at making sense of ARM assembly code (which you'll have to
generate on your own using the methods talked about on here over the past
week), it would be very, very difficult. Don't let that stop you.

I've noticed that the Newton ROM board (which uses out of production Sharp
Flash-ROMs) at least looks like a glorified Sharp copier ROM board. It
even uses the same snap in retainer, but the board itself has added height
to accomododate more chips. Only once have I seen a different chip
layout, however. I contacted a guy selling Sharp Flash programmers that
actually fit these, but he told me that there's no real consistancy with
Sharp's ROM board design and that it only works with a slim range. If we
knew more about the chips, we could ideally build an adapter to read/write
and even support added chips (I would love to see added NewtonOS APIs, let
alone Linux). Or better yet, if anyone has an office Sharp copier on the
fritz, see if you can get a peek at the chip numbering should it have a
module-based ROM board. Better yet, yank it out, scan it, and send it my
way so I can compare wiring layout.

Throughout MessagePad development, there's always been close
intercommunication between the ARM processor and a control ASIC as well.
In the 2x00 and I believe the eMate, the main control chip is a Cirrus
Logic PS7010. I'll let everyone know how things go if I can schedule time
with a guy who has control over an electron microscope, but these chips
are multi-layer and technical sheets are what we desperately need. I
believe the PS7010 may be based on an ARM 610 or 710 core, but this is
speculation.

There are currently a number of Linux projects based on ARM designs,
which include info on boatloading and such (Itsy, iPaq, Pleb, LART, and
others at handhelds.org). There's also Linux support for odd things, such
as the keyboard. You can also request a free evaluation CD of the "ARM
Developer Suite" for Windows as well as their technical documents CD.

If you know anything about python programming--or better yet, converting
python code to C or C++ :) --take a look at the abandoned
gnuton.sourceforge.net project. But also check out sites like:

http://www.kallisys.com/newton/bowels/?lg=en
http://www.geocities.com/newtontechspecs
http://www.pda-soft.de/hardware.html
        (some links need to be fixed)
http://www.chemia.polsl.gliwice.pl/~philz
        (needs to fix access permissions)

-Jeff

-- 
Read the List FAQ/Etiquette: http://www.newtontalk.net/faq.html
Read the Newton FAQ: http://www.chuma.org/newton/faq/
This is the NewtonTalk mailing list - http://www.newtontalk.net/



This archive was generated by hypermail 2.1.2 : Thu Oct 31 2002 - 12:02:55 EST